Minus Zero unveils AI-powered autopilot system for Indian roads

Autonomous mobility startup Minus Zero has launched an end-to-end autopilot system designed specifically for complex Indian traffic conditions. The vision-based solution, developed using foundational AI models, enables navigation on unstructured roads with limited or no lane markings and can handle unpredictable obstacles like animals, pushcarts, and two-wheelers.

The Bengaluru-based company has taken a self-supervised learning approach, moving away from conventional rule-based algorithms. This allows the system to train itself using large volumes of raw data without relying on human-labeled inputs.

Minus Zero’s autopilot system also eliminates the need for expensive LiDAR sensors and HD maps, using a minimal hardware vision-first stack that’s more economical and easier to deploy.

Tested across challenging routes in Bengaluru, the system marks the first time foundational models have been applied to Indian roads for autonomous driving. Capable of adapting to unseen environments and new obstacles, the platform is built to scale rapidly across geographies and vehicle formats.

Minus Zero is working with multiple OEMs to bring the system to market within two years. This move positions the company at the forefront of India’s emerging autonomous driving sector as it builds globally competitive deeptech solutions from the ground up.